EP2270458A1 - Method and device for detecting leaks - Google Patents

Abstract

The method involves sucking out gas, detecting test gas in the sucked out gas and generating a quantity signal (MS) corresponding to the measured leakage rate, processing the quantity signal by reducing the quantity signal if a zero signal is produced, determining the signal noise, determining a lower display threshold (AG) using the signal noise and only indicating a leak if the quantity signal exceeds the signal noise. An independent claim is also included for a leak location device.

Description

Procedure for leak detection

The invention relates to a method for leak detection, wherein gas is sucked out and the presence of test gas in the extracted gas is detected.

The smallest leak rates of a container can be reliably detected by the vacuum method. The smaller the leak rate, the greater the requirements for cleanliness and final vacuum. In a local leak detection, the container is evacuated with a vacuum pump until the required test pressure for the leak detector is reached. Leakage points of the container are then sprayed on the outside with a fine test gas jet. In the container penetrating test gas is pumped out of the vacuum pumping device and detected by a mass spectrometer.

Helium is often used as the test gas for leak detectors. One difficulty is the limited selectivity of the mass spectrometer. On containers for the leak measurement water settles both outside and inside. The H ₂ component of water also includes portions whose presence is sensitive to the measurement of helium. At the beginning of the pumping off, the quantity signal, which is supposed to represent only the amount of the test gas, is superimposed on a disturbance variable which is produced by the presence of water or other impurities. The disturbance decays as the pumping time increases to asymptotically approach a horizontal line. However, this line never reaches the value zero because an absolute vacuum can be achieved just as little as an absolute freedom from leaks. It is therefore a matter of the particular application which range of the decaying mass flow curve is selected for the leak measurement.

Since the curve of the quantity signal decays with increasing pumping time, the sudden occurrence of test gas, which causes an increase of the curve of the quantity signal, is superimposed by the decaying background signal. When the background signal decreases to a greater extent than the detection signal builds up, a detection signal is not detected at all. The corresponding leak is not detected by the leak detector; it remains invisible.

It has already been proposed to provide a leak detector with a zero function. The device is provided with a zero-key that can be pressed by the user to subtract from the current signal the previously occurred signal. In this way, the signal amplitude is set to zero. Subsequently, as the subsiding subsonic signal decreases to a greater extent as the quantity signal increases due to leakage, a negative signal results in which the signal rise caused by the leak is neither detectable nor measurable.

In the hitherto conventional methods, the user can press the zero button at any time, with the background signal being set to zero. The result is that a seemingly low leak rate is displayed while the actual leak rate is higher. Such an operation can have fatal consequences. It is important that no leakage rate is hidden. On the other hand, a too much indicated leak rate is less problematic.

The invention has for its object to provide a method for leak detection, in which the leak detection reliability is increased.

A first variant of the method according to the invention is specified in claim 1 and a second variant is specified in claim 6. Both variants have in common that the signal unrest of the measurement signal is determined and evaluated. Signal unrest is the change in the signal background per unit of time. In the first variant of the method, the quantity signal is not set to zero by the generation of the command signal, which is usually referred to as zero signal, but only reduced to a lower display limit, so that the signal value obtained is still positive. The lower display limit indicates how far a leak can be detected. There is no blocking of the function of the zero signal. If the detected leak rate is above the lower display limit, it will be displayed; otherwise not. The method does not work with a zero level. It is determined to which signal unrest which leak rate is still sufficiently measurable. It is thus automatically indicated, which sensitivity the device has at the time when the zero signal has been generated by pressing the zero button.

In the second variant of the method according to the invention, there is likewise a determination of the signal unrest of the quantity signal. The user specifies a so-called "trigger value", which indicates the desired sensitivity of the leak detection, that is, for example, a leak rate value of 10 ^-10 mbar I / s. The release of the Zero function is only possible if the signal unrest of the quantity signal is less than the trigger value. As long as the instability of the volume signal is greater than the trigger value, the zero function is disabled and there is no leak indication. The user must therefore wait until the quantity signal has calmed down until the desired sensitivity set using the trigger value has been reached.

According to the invention, the zero function is disabled for the user of the device. The release of the Zero function occurs only when the device can reliably measure leak rates that correspond to the entered trigger value as a result of signal saturation of the volume signal.

The invention further relates to corresponding leak detectors for the first and the second variant of the method.

In the following, embodiments of the invention will be explained in more detail with reference to the drawings.

Figur 1

eine Prinzipzeichnung eines Lecksuchgeräts, das nach dem erfindungsgemäßen Verfahren arbeitet,

Figur 2

ein Blockdiagramm der ersten Variante des Verfahrens,

Figur 3

ein Zeitdiagramm des Mengensignals bei der ersten Variante,

Figur 4

ein Blockdiagramm der zweiten Variante, und

Figur 5

ein Zeitdiagramm des Mengensignals bei der zweiten Variante.

Show it:

FIG. 1

a schematic drawing of a leak detector, which operates according to the inventive method,

FIG. 2

a block diagram of the first variant of the method,

FIG. 3

a time diagram of the quantity signal in the first variant,

FIG. 4

a block diagram of the second variant, and

FIG. 5

a timing diagram of the quantity signal in the second variant.

The leak detector 10 after FIG. 1 has an inlet flange 11 to which the container to be tested is connected. From the inlet flange 11, a line 12 leads to the vacuum pump device 13. This consists of a turbomolecular pump 14 and one of these downstream backing pump 15. The line 12 is connected to a side inlet of the turbomolecular pump 14 whose inlet side is connected to a mass spectrometer 16. In the turbomolecular pump 14, the test gas contained in the sucked gas, for example, helium, enters the mass spectrometer 16 in countercurrent to the conveying direction to be identified. The mass spectrometer 16 supplies the quantity signal representing the recognized amount of test gas to a microcomputer 17 which executes the processing explained below. The microcomputer is connected to an operating unit 18, which contains a display device 19, for example a screen, an input device 20 with different keys and a zero key 21.

To the inlet flange 11 of the container to be tested 25 is connected, which has an (unwanted) leak 26. The leak is leaked with test gas, e.g. Helium sprayed from a spray gun 27. The test gas, which enters the container 25, passes via the turbomolecular pump 14 to the mass spectrometer 16. The amount of the test gas is displayed as a quantity signal MS on the display device 19 in the form of a curve and / or as a numerical value.

At the beginning of the suction, the quantity signal MS has FIG. 3 a relatively high value. During aspiration, the volume signal MS sounds asymptotic. The high value of the volume signal is due to water and other impurities, as well as residual helium contained in the extracted gas. The quantity signal MS thus has a foreign-influenced drift, whereby the external influences can far exceed the measurement signal. The representation of the quantity signal MS along the coordinate in FIG. 3 is constructed logarithmically by powers of ten. The quantity signal MS shown in the initial region of the curve results without the Presence of a leak. It forms the signal background on which a leak is still to be detected.

FIG. 2 shows the structure of an embodiment of the evaluation circuit, or the processing in the microcomputer 17. In this case, the quantity signal MS, which is to indicate the leak rate Q, is supplied via a subtractor to the one input of a maximum value selector 30. The subtractor 31 receives in its subtraction input the signal of a memory 32 which is connected to the output of the subtracter 31. The memory is activated by a "zero" signal on line 33 in such a way that it initiates a subtraction process in which the output signal of the subtractor 31 is subtracted from the quantity signal MS. This results in the reduced quantity signal MS _z , which is the one input of the maximum value selector 30 is supplied. In the subtracter 31, the total magnitude of the quantity signal is subtracted, giving the value MS _z of zero.

der Signalunruhe bildet. Dieses Signal ist die Ableitung der Leckrate nach der Zeit. Es ist umso größer, je steiler der Abfall des Mengensignals MS ist (Figur 3). Das Signal der Differentialschaltung 34 wird mit einer Konstanten 1/k multipliziert. Daraus wird die untere Anzeigegrenze AG ermittelt. Der Wert von AG wird dem zweiten Eingang des Maximalwertselektors 30 zugeführt. Aus den beiden Eingangsgrößen MS_z und AG wählt der Maximalwertselektor den größten aus. Dieser wird an der Anzeigevorrichtung als Anzeigesignal AS zur Anzeige gebracht.The quantity signal MS is also supplied to a differential circuit 34 which is a signal from the leak rate Q. $\frac{\mathit{dQ}}{\mathit{dt}}$

the signal disturbance forms. This signal is the derivative of the leak rate over time. It is all the greater, the steeper the drop of the quantity signal MS is ( FIG. 3 ). The signal of the differential circuit 34 is multiplied by a constant 1 / k. From this, the lower display limit AG is determined. The value of AG is supplied to the second input of the maximum value selector 30. From the two input variables MS _z and AG, the maximum value selector selects the largest one. This is displayed on the display device as a display signal AS for display.

FIG. 3 shows the course of the display signal AS, so the history with pressing the zero button. In FIG. 3 Assume that at time t _1, the zero button 21 is pressed to prepare for leak detection. Shortly thereafter will with the spray gun 27 test gas sprayed against the container 25. The spraying takes place at time t ₂ and ends at time t ₃ .

It is appreciated that the indication signal generated by the maximum value selector 30 AS at time t ₁ decreases to the value of the indication limit AG, because of the time t ₁ to the display limit AG is greater than the then through the subtractor 31 generated volume signal MS _z. If at the time t ₂ a spraying of the leak occurs, the measurement signal increases beyond the display limit AG by spraying, so that a pulse 37 is formed, which, however, already decays during spraying according to the general drift and finally back on the curve of the display limit AG ends. The pulse 37 is clearly identifiable within the curve of the indication signal AS and thus recognizable as a leak.

After some time, the process of activating the Zero button can be repeated to then spray the container again. There is a determination and display of the lower display limit AG, which is displayed on the display device. A leak is indicated by the change of the indicated value of AG to a higher value, corresponding to the pulse 37. On the display device so the lower display limit AG is constantly indicated, in which a leak rate is still sufficiently well displayed. The user can thus perform a leak detection with the previously displayed sensitivity when displaying a desired display limit AG.

FIG. 4 shows an embodiment of the second variant of the invention. The measured leak rate Q is also supplied here as a quantity signal MS to a subtractor 31. The output of the subtractor is connected via a memory 32 to the subtraction input of the subtractor. The signal output from the memory 32 is caused by a signal on a line 33. The zero signal inputted to a zero key is applied to a lock 40 which receives the Signal for line 33 generated. The barrier 40 is activated by a signal on a line 41.

At a trigger input 42, which is the input device 20 of FIG FIG. 1 the user enters a trigger value T in mbar * l / s. The trigger value initially represents a limit whose exceeding indicates a "too large" leak.

The value Q _z , which the measured leak rate has assumed after activating the zero function, is fed to the one input of a comparator 43. The other input of the comparator receives the trigger value T. The comparator 43 produces an output signal when Q _z > T. This output signal activates the trigger alarm 44, which indicates that the quantity signal is greater than the entered trigger value and thus exceeds the limit value. This means the detection of a leak. The size of the leak is displayed on the leak rate indicator 45, which receives the Q _z signal.

bildet. Das Ausgangssignal D der Differentialschaltung 50 gibt die Signalunruhe der Leckrate Q an, also die (negative) Steigung des zeitlichen Verlaufs des Mengensignals. Das Signal D wird zwei Logikschaltungen 51 und 52 zugeführt. Die erste Logikschaltung 51 liefert ein Ausgangssignal, wenn folgende Bedingung erfüllt ist $$\left(\mathrm{Q}\mathrm{<}\mathrm{0}\right)\mathrm{\wedge }\left(\left|\mathrm{D}\right|\mathrm{>}\mathrm{c}\mathrm{*}\mathrm{T}\right)\mathrm{.}$$

The quantity signal MS, which represents the leak rate Q, is supplied to a differential circuit 50, which outputs the differential quotient $\frac{\mathit{dQ}}{\mathit{dt}}$

forms. The output signal D of the differential circuit 50 indicates the signal disturbance of the leak rate Q, ie the (negative) slope of the time profile of the quantity signal. The signal D is supplied to two logic circuits 51 and 52. The first logic circuit 51 provides an output signal when the following condition is satisfied $$\left(\mathrm{Q}\mathrm{<}\mathrm{0}\right)\mathrm{\wedge }\left(\left|\mathrm{D}\right|\mathrm{>}\mathrm{c}\mathrm{*}\mathrm{T}\right)\mathrm{,}$$

The second logic circuit 52 generates an output signal when the following condition is satisfied $$\left(\mathrm{D}\mathrm{>}\mathrm{0}\right)\vee \left(\left|\mathrm{D}\right|<\mathrm{k}\mathrm{*}\mathrm{T}\right)\mathrm{,}$$

Here is
D: the signal disturbance in mbar * l / s per minute
T: the set trigger value in mbar * l / s
c, k: constants with c> k.

By selecting the constants c and k, the minimum duration that a leak of size T is visible to the user can be specified, before it becomes "invisible" again due to the negative drift of the leak rate signal. It must be c> k, so that a hysteresis between "lock" and "release" of the zero function is given. This feature prevents zero operation of the zero function and ensures that leaks the size of the set trigger value are detected by the user.

The output signals of the logic circuit 51 and 52 control a flip-flop 53, at whose output the line 41 is connected, which controls the lock 40. The signal of the logic circuit 51 controls the set input S and the signal of the logic circuit 52 controls the reset input R of the flip-flop 53. The output of the flip-flop is connected to a display device 54 having two differently colored lights 55,56. The lamp 55 lights up when the Zero function is enabled and the lamp 56 lights up when the Zero function is locked.

ebenfalls in Figur 5 eingetragen ist. Im Punkt P schneidet die Kurve $-\frac{D}{k}$

den Trigger-Wert T. Dies bedeutet, dass durch die Sperre 40 in Figur 4 die Zero-Funktion freigegeben wird. Die Anzeigeleuchte 55 leuchtet. Der Benutzer kann nun manuell das Zero-Signal erzeugen. Vom Zeitpunkt t₁ bis zum Zeitpunkt t₂ sprüht der Benutzer mit der Sprühpistole 27 Testgas gegen das Leck. Dies führt zu einem positiven Anstieg 60 des Mengensignals. Vom Zeitpunkt t₂ an wird die Zero-Funktion gesperrt. Sobald das Signal D wieder den eingestellten Trigger-Wert T von oben nach unten passiert, wird vom Zeitpunkt t₃ die Zero-Funktion wieder freigegeben. In Figur 5 ist die Hysterese, die durch die Konstanten c und k hervorgerufen wird, aus Gründen der Übersichtlichkeit nicht dargestellt. FIG. 5 shows the timing of the leak rate Q, which forms the quantity signal MS. The representation of the quantity signal MS is also in FIG. 5 logarithmically built up by powers of ten. From the leak rate Q, the signal disturbance D is determined, which scales as a curve $-\frac{D}{k}$

also in FIG. 5 is registered. in the Point P intersects the curve $-\frac{D}{k}$

the trigger value T. This means that through the lock 40 in FIG. 4 the zero function is released. The indicator light 55 lights up. The user can now manually generate the zero signal. From time t ₁ to time t ₂ , the user sprays test gas against the leak with the spray gun 27. This leads to a positive increase 60 of the quantity signal. From time t ₂ on, the zero function is disabled. As soon as the signal D again passes the set trigger value T from top to bottom, the zero function is released again from the time t ₃ . In FIG. 5 is the hysteresis, which is caused by the constants c and k, not shown for reasons of clarity.

Signal rise 60 indicates a detected leak.

The variant after the FIGS. 4 and 5 based on the fact that it is determined whether the leak rate in the amount of the desired trigger value T is still measurable. If it can be measured, the zero function is released; if it is not measurable, the zero function is blocked.

In the embodiments described above, each gas is sucked from a container to check this gas for test gas. Deviating from this, the invention is also applicable to sniffer leak detection, in which a leak is detected by the fact that a sucking sniffer probe draws in ambient air at the point to be tested.

Claims (5)

Procedure for leak detection with the following steps: - suction of gas, Detecting test gas in the extracted gas, and generating a quantity signal (MS) corresponding to the measured leak rate (Q), - Determination of the signal unrest $\left(\frac{\mathit{dQ}}{\mathit{dt}}\right)$

of the quantity signal as
Derivation of the measured leak rate after the time,
characterized by the further steps Processing of the quantity signal (MS) in such a way that a reduction of the quantity signal takes place when a zero signal is generated, Determining a lower display limit (AG) on the basis of the determined signal unrest, - Display of a leak only if the quantity signal (MS) exceeds the lower display limit (AG). A leak detection method according to claim 1, wherein the detection of signal perturbation $\left(\frac{\mathit{dQ}}{\mathit{dt}}\right)$

by differential formation of the quantity signal (MS). A leak detection method according to claim 1 or 2, wherein the signal perturbation is divided by a quotient (k) to obtain the display limit (AG). Leak detection method according to one of Claims 1 to 3, in which the maximum of leakage rate (Q) and indication limit (AG) is supplied to a leak rate indicator. A leak detector comprising a vacuum port (11), a high vacuum pumping device (13) connected to the vacuum port, a check gas sensor (16) providing a quantity signal (MS), and a differential circuit (34) generating a signal perturbation value from the quantity signal,
characterized,
in that the leak detector comprises a microcomputer (17) and an operating device (18) which contains a display device (19) and a zero signal input device (21) for reducing the quantity signal supplied by the test gas sensor,
characterized in that a lower display limit (AG) is determined from the signal unevenness value, and that the display device (19) contains a maximum value selector (30) which selects a maximum value from the reduced quantity signal (MS) and the display limit (AG), wherein the Display device (19) displays a deviating from the display limit leak rate only if the reduced quantity signal (MS) exceeds the display limit (AG).

Images